ORIGINAL

[sig fs] = read_NIST_file(filename); % read the input file, assume filename

is supplied
NFFT = 1024;
WINLEN = 0.025; %frame length in ms
WINSTEP = 0.01;
frames = frame_sig(sig, WINLEN*fs, WINSTEP*fs, @hamming);
cspec = fft(frames,NFFT,2); % complex spectrum
pspec = abs(cspec).^2; % power spectrum of noisy signal
phase = angle(cspec);

% do spectral subtraction, produce modified_pspec

noise_est = mean(pspec(1:3,:)); % noise_est is estimated from first three
frames
clean_spec = pspec - repmat(noise_est,size(pspec,1),1); % subtract noise_est
from pspec
clean_spec(clean_spec < 0) = 0; % negative power spectrum is not allowed

reconstructed_frames = ifft(sqrt(clean_spec).*exp(phase),NFFT,2);
reconstructed_frames = real(reconstructed_frames(:,1:WINLEN*fs)); %
sometimes small complex residuals stay behind
enhanced_signal =
deframe_sig(reconstructed_frames,length(sig),WINLEN*fs,WINSTEP*fs,@hamming);

soundsc(enhanced_signal,fs); % listen to our enhanced signal

Traducido
[sig fs] = read_NIST_file (nombre de archivo); % lee el archivo de entrada,
asume que se proporciona el nombre de archivo
NFFT = 1024 ;
WINLEN = 0.025 ; % de longitud de trama en ms
WINSTEP = 0.01 ;
marcos = frame_sig (sig, WINLEN * fs, WINSTEP * fs, @hamming);
cspec = fft (cuadros, NFFT, 2 ); % espectro complejo
pspec = abs (cspec). ^ 2; % de espectro de potencia de señal ruidosa
fase = ángulo (cspec);

% hacer sustracción espectral, producir modified_pspec

noise_est = mean (pspec ( 1 : 3 , :)); % noise_est se estima a partir de los
primeros tres cuadros
clean_spec = pspec - repmat (noise_est, size (pspec, 1 ), 1 ); % sustraer
noise_est de pspec
clean_spec (clean_spec < 0 ) = 0 ; % de espectro de potencia negativo no
está permitido

reconstructed_frames = ifft (sqrt (clean_spec). * exp (phase), NFFT, 2 );

reconstructed_frames = real (reconstructed_frames (:, 1 : WINLEN * fs)); % a
veces quedan pequeños residuos complejos
Enhanced_signal = deframe_sig (reconstructed_frames, length (sig), WINLEN *
fs, WINSTEP * fs, @ hamming);

sonidosc (señal mejorada, fs); % escucha nuestra señal mejorada

CÓDIGO DE MATLAB RESTA ESPECTRAL

fprintf('--- Audio Signal Denoising using Spectral Subtraction ---\n\n');

%load noise sound example

[y,Fe]=audioread('sample.wav');

x=y(100000:end,1).'; %remove the beginning of the sample

Nx=length(x);

%algorithm parameters

apriori_SNR=1; %select 0 for aposteriori SNR estimation and 1 for apriori

alpha=0.05; %only used if apriori_SNR=1

beta1=0.5;

beta2=1;

lambda=3;

%STFT parameters
NFFT=1024;

window_length=round(0.031*Fe);

window=hamming(window_length);

window = window(:);

overlap=floor(0.45*window_length); %number of windows samples without

overlapping

%Signal parameters

t_min=0.4; %interval for learning the noise

t_max=1.00; %spectrum (in second)

%construct spectrogram

[S,F,T] = spectrogram(x+i*eps,window,window_length-overlap,NFFT,Fe); %put a

short imaginary part to obtain two-sided spectrogram

[Nf,Nw]=size(S);

%Noisy spectrum extraction

t_index=find(T>t_min & T<t_max);

absS_noise=abs(S(:,t_index)).^2;

noise_spectrum=mean(absS_noise,2); %average spectrum of the noise

noise_specgram=repmat(noise_spectrum,1,Nw);

%Estimate SNR

absS=abs(S).^2;
SNR_est=max((absS./noise_specgram)-1,0);

if apriori_SNR==1

SNR_est=filter((1-alpha),[1 -alpha],SNR_est);

end

%Compute Attenuation Map

an_lk=max((1-lambda*((1./(SNR_est+1)).^beta1)).^beta2,0);

STFT=an_lk.*S;

%Compute Inverse STFT

ind=mod((1:window_length)-1,Nf)+1;

output_signal=zeros((Nw-1)*overlap+window_length,1);

for indice=1:Nw %Overlapp add technique

left_index=((indice-1)*overlap) ;

index=left_index+[1:window_length];

temp_ifft=real(ifft(STFT(:,indice),NFFT));

output_signal(index)= output_signal(index)+temp_ifft(ind).*window;

end

%----------------- Display Figure ------------------------------------

%show temporal signals

figure

subplot(2,1,1);

t_index=find(T>t_min & T<t_max);

plot([1:length(x)]/Fe,x);

xlabel('Time (s)');

ylabel('Amplitude');

hold on;

noise_interval=floor([T(t_index(1))*Fe:T(t_index(end))*Fe]);

plot(noise_interval/Fe,x(noise_interval),'r');

hold off;

legend('Original signal','Noise Only');

title('Original Sound');

%show denoised signal

subplot(2,1,2);

plot([1:length(output_signal)]/Fe,output_signal );

xlabel('Time (s)');

ylabel('Amplitude');

title('Sound without noise');

%show spectrogram

t_epsilon=0.001;

figure

S_one_sided=max(S(1:length(F)/2,:),t_epsilon); %keep only the positive

frequency

pcolor(T,F(1:end/2),10*log10(abs(S_one_sided)));

shading interp;

colormap('hot');
 title('Spectrogram: speech + noise');

xlabel('Time (s)');

ylabel('Frequency (Hz)');

figure

S_one_sided=max(STFT(1:length(F)/2,:),t_epsilon); %keep only the positive

frequency

pcolor(T,F(1:end/2),10*log10(abs(S_one_sided)));

shading interp;

colormap('hot');

title('Spectrogram: speech only');

xlabel('Time (s)');

ylabel('Frequency (Hz)');

%Listen results

fprintf('\nPlay 5 seconds of the Original Sound:');

audioplayer(x(1:5*Fe),Fe);

fprintf(' OK\n');

fprintf('Play 5 seconds of the new Sound: ');

audioplayer(output_signal(1:5*Fe),Fe);

fprintf('OK\n');

fprintf('Write denoised_speech.wav');

audiowrite('denoised_signal.wav',output_signal,Fe);

 © 2019 GitHub, Inc.

 Terms
 Privacy
 Security
 Status
 Help

 Contact GitHub
 Pricing
 API
 Training
 Blog
 About